- Wine-Devel - list.winehq.org

server: Fix for possible desktop variable being NULL when set_process_default_desktop is called
by Dan Kegel April 7, 2006

April 7, 2006

Mike Hearn wrote: >I'm not really convinced a static >analysis can catch things that we "know" cannot happen (eg it'd be a >violation of the X protocol or something) and so rely on them without >explicit checks. To be secure, I think we have to handle X protocol violations, too. - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

1 0

Compiling wine 0.9.11 under solaris 10
by Chiaki April 7, 2006

April 7, 2006

I am trying to build wine-0.9.11 on a solaris express system (nevada b29). But at the end of compilation, I got link errors in a couple places. (messages are given below.) I followed the guidance at http://www.blastwave.org/wine/winemaking.html I had not known the tips mentioned at http://wiki.jswindle.com/index.php/Installing_Wine#Solaris before I tried compilation, Because I didn't know, instead of disabling some macros as suggested in the above URL, I hacked the source files so that the previously ignored files were compiled under Solaris 10 (or Solaris 11 beta). (I listed some of these into a mail list mentioned at blastware web page [solaris package repository] http://lists.blastwave.org/pipermail/users/2006-April/thread.html when I was tinkering with 0.9.10. The changes are basically the same for the case of 0.9.11 with a few additions now that I have configured 0.9.11 in a slightly different manner from the way 0.9.10 was configured. And the link errors I saw have not changed between 0.9.10 and 0.9.11) Anyway, at the end of compilation, I got link errors in a couple places. ../libs/wine/libwine.so: undefined reference to `write(a)SUNW_0.9' ../libs/wine/libwine.so: undefined reference to `read(a)SUNW_0.9' collect2: ld returned 1 exit status These seem to be due to versioning(?) of read/write system call symbols or something. (SUNW seems to be symbol used by SUN to mark symbols coming from their libraries, etc.. It is market symbol for Sun Microsystems. Most of Sun-supplied `packages' for solaris is named like SUNWxyz to denote that they are coming from Sun instead of 3rd parties.) Have anybody seen these link problems before? Are there any workarounds? Or am I using incorrect ld/gas, etc? Tool chain information. $ uname -a SunOS solaris-nv-28-home 5.11 snv_28 i86pc i386 i86pc $ which gcc /usr/sfw/bin/gcc bash-3.00$ gcc --version gcc (GCC) 3.4.3 (csl-sol210-3_4-20050802) Copyright (C) 2004 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. (I have replaced /usr/ccs/bin/ld invoked with the above to be gnu ld instead of the original sun ld.) bash-3.00$ /usr/ccs/bin/ld --version GNU ld version 2.16.1 Copyright 2005 Free Software Foundation, Inc. This program is free software; you may redistribute it under the terms of the GNU General Public License. This program has absolutely no warranty. bash-3.00$ The eerors at this stage are only seen for two programs namely wine-kthread and wine-server at this moment. The entire log for the last remaining errors follows here. (Line folding is done by the mailer here, sorry about this.) cd /home/ishikawa/local-src/wine-sources/wine-0.9.11/ bash-3.00$ make -k gcc -I/opt/csw/include/freetype2 -L/opt/csw/lib/ -o wine-kthread -Wl,--export -dynamic -Wl,--section-start,.interp=0x7bf00400 -Wl,--rpath,\$ORIGIN/`../tools /relpath /usr/local/bin /usr/local/lib` kthread.o main.o -L../libs/wine -lwine -L../libs/port -lwine_port ../libs/wine/libwine.so: undefined reference to `write(a)SUNW_0.9' ../libs/wine/libwine.so: undefined reference to `read(a)SUNW_0.9' collect2: ld returned 1 exit status *** Error code 1 make: Warning: Target `all' not remade because of errors Current working directory /home/ishikawa/local-src/wine-sources/wine-0.9.11/lo ader *** Error code 1 The following command caused the error: cd loader && make `libdinput.def.a' is up to date. ../../tools/winegcc/winegcc -B../../tools/winebuild -mwindows curses.o dialog .o registry.o user.o wineconsole.o wineconsole_res.res -o wineconsole.exe. so -Wl,--rpath,\$ORIGIN/`../../tools/relpath /usr/local/lib/wine /usr/local/li b` -L../../dlls -lcomctl32 -luser32 -lgdi32 -ladvapi32 -lkernel32 -lntdll -L.. /../libs -lwine -L../../libs/port -lwine_port -lrt -lthread -lsocket -lnsl - lcurses -Wb,-dcomctl32 -Wb,-duser32 -Wb,-dgdi32 gcc -I/opt/csw/include/freetype2 -L/opt/csw/lib/ -o wineserver -Wl,--rpath,\$ ORIGIN/`../tools/relpath /usr/local/bin /usr/local/lib` atom.o change.o clas s.o clipboard.o console.o context_alpha.o context_i386.o context_powerpc. o context_sparc.o context_x86_64.o debugger.o directory.o event.o fd.o file.o handle.o hook.o mailslot.o main.o mapping.o mutex.o named_pipe.o object.o process.o ptrace.o queue.o region.o registry.o request.o sem aphore.o serial.o signal.o snapshot.o sock.o symlink.o thread.o timer.o token.o trace.o unicode.o user.o window.o winstation.o sunptrace.o -L. ./libs/wine -lwine -L../libs/unicode -lwine_unicode -L../libs/port -lwine_port -lrt -lthread -lsocket -lnsl ../libs/wine/libwine.so: undefined reference to `write(a)SUNW_0.9' ../libs/wine/libwine.so: undefined reference to `read(a)SUNW_0.9' collect2: ld returned 1 exit status *** Error code 1 make: Warning: Target `all' not remade because of errors Current working directory /home/ishikawa/local-src/wine-sources/wine-0.9.11/se rver *** Error code 1 The following command caused the error: cd server && make make: Warning: Target `all' not remade because of errors bash-3.00$ TIA -- int main(void){int j=2006;/*(c)2006 cishikawa. */ char t[] ="<CI> @abcdefghijklmnopqrstuvwxyz.,\n\""; char *i ="d>qtCIuqivb,gCwe\np(a).ietCIuqi\"tqkvv is>dnamz"; while(*i)((j+=(int)strchr(t,*i++)-(int)t),(j%=sizeof t-1), (putchar(t[j])));return 0;}/* under GPL */

1 0

Re: [dsound] set default primary buffer sample rate and bits per sample
by Alexandre Julliard April 7, 2006

April 7, 2006

Robert Reif <reif(a)earthlink.net> writes: > Programs that are written specifically for 2000 and xp don't > bother to set the primary buffer format because it's a noop. However > wine is patterned after win9x and DirectX 7 or > earlier which has a real primary buffer and expects the > program to change the primary buffer format if necessary. > > This patch provides a way to change the primary buffer format > from 22050x8 to an application specific value or a system wide > value for programs that don't. Shouldn't we make it work the win2k way then, instead of requiring users to configure things manually? -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: server: Fix for possible desktop variable being NULL when set_process_default_desktop is called
by Alexandre Julliard April 7, 2006

April 7, 2006

Robert Shearman <rob(a)codeweavers.com> writes: > ChangeLog: > The desktop variable could be NULL when set_process_default_desktop is > called, so check for this. (Found by Coverity). I don't see how this could happen. Did I miss something? -- Alexandre Julliard julliard(a)winehq.org

3 6

Re: buffer check in comdlg
by Robert Shearman April 7, 2006

April 7, 2006

Aric Stewart wrote: > if(fodInfos->unicode) > { > LPOPENFILENAMEW ofn = fodInfos->ofnInfos; >- strcpyW( ofn->lpstrFile, lpstrPathSpec); >- memcpy( ofn->lpstrFile + nSizePath, lpstrFileList, sizeUsed*sizeof(WCHAR) ); >+ if (nSizePath+sizeUsed > ofn->nMaxFile) >+ { >+ *((WORD*)ofn->lpstrFile)=nSizePath+sizeUsed; >+ FILEDLG95_Clean(hwnd); >+ EndDialog(hwnd,FALSE); >+ COMDLG32_SetCommDlgExtendedError(FNERR_BUFFERTOOSMALL); >+ return FALSE; >+ } >+ else >+ { >+ strcpyW( ofn->lpstrFile, lpstrPathSpec); >+ memcpy( ofn->lpstrFile + nSizePath, lpstrFileList, sizeUsed*sizeof(WCHAR) ); >+ } > } > > This formatting doesn't look at all consistent, either with itself or the other code there. -- Rob Shearman

2 1

WineD3D: Add a WINEMAKEFOURCC macro
by Stefan Dösinger April 7, 2006

April 7, 2006

Hello, Alexandre, if you think > that it should be in the wined3d headers, you can add > #define MAKEFOURCC(a,b,c,d) ((a << 0) | (b << 8) | (c << 16) | (d << 24)) > in line 169 in wined3d_interface.h, or ask me to send a patch for that. Forget about that, I forgot that wined3d_types is included before that line. Attached is a patch which creates WINEMAKEFOURCC and uses this. Changelog: Stefan Dösinger: Add a WINEMAKEFOURCC macro to wined3d_types.h

1 0

Re: advpack: Rewrite get_parameter to be able to handle an empty parameter
by James Hawkins April 7, 2006

April 7, 2006

On 4/7/06, James Hawkins <truiken(a)gmail.com> wrote: > Hi, > > Changelog: > * Rewrite get_parameter to be able to handle an empty parameter. > blah Ignore this, because I sent too much in one patch. -- James Hawkins

1 0

new to wine, Guild Wars don't display correctly
by BONNEL Christophe April 7, 2006

April 7, 2006

3 6

Re: msi: Use a debug channel to turn on transform debugging. (Coverity)
by Alexandre Julliard April 7, 2006

April 7, 2006

Mike McCormack <mike(a)codeweavers.com> writes: > --- a/dlls/msi/table.c > +++ b/dlls/msi/table.c > @@ -1604,13 +1604,12 @@ static MSIRECORD *msi_get_transform_reco > USHORT mask = *rawdata++; > MSICOLUMNINFO *columns = tv->columns; > MSIRECORD *rec; > - const int debug_transform = 0; > > rec = MSI_CreateRecord( tv->num_cols ); > if( !rec ) > return rec; > > - if( debug_transform ) MESSAGE("row -> "); > + if( TRACE_ON(msidb) ) MESSAGE("row -> "); You should use TRACE instead of MESSAGE then, it will be a lot more readable. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: x11drv: (bug 4685) properly handle VK_LMENU input (resend)
by Dmitry Timoshkov April 7, 2006

April 7, 2006

<tkho(a)ucla.edu> wrote: A couple of suggestions about coding style: > +struct transition_s { > + BYTE wVk; It would be more logical to use WORD for VK codes as everywhere else. > + BYTE before_state; > + BYTE _todo_wine; > +}; Same for _todo_wine, it should be BOOL. > + > +struct sendinput_test_s { > + WORD wVk; > + DWORD dwFlags; > + struct transition_s expected_transitions[MAXKEYEVENTS+1]; > + UINT expected_messages[MAXKEYMESSAGES+1]; > +} sendinput_test[] = { > + /* test ALT+F */ > + {VK_LMENU, 0, {{0x12, 0x00, 0}, {0xa4, 0x00, 0}, {0}}, {0x104, 0}}, > + {'F', 0, {{0x46, 0x00, 0}, {0}}, {0x104, 0x106, 0x112, 0}}, > + {'F', KEYEVENTF_KEYUP, {{0x46, 0x80, 0}, {0}}, {0x105, 0}}, > + {VK_LMENU, KEYEVENTF_KEYUP, {{0x12, 0x80, 0}, {0xa4, 0x80, 0}, {0}}, > + {0x101, 0}}, Please use VK names and symbolic message names in the above arrays to make them more readable. > /*********************************************************************** > * X11DRV_send_keyboard_input > @@ -1128,14 +1141,16 @@ > { > message = WM_KEYUP; > if ((key_state_table[VK_MENU] & 0x80) && > - ((wVk == VK_MENU) || (wVk == VK_CONTROL) || !(key_state_table[VK_CONTROL] & 0x80))) > + ((VkStripLR(wVk) == VK_MENU) || (VkStripLR(wVk) == VK_CONTROL) > + || !(key_state_table[VK_CONTROL] & 0x80))) > { > if( TrackSysKey == VK_MENU || /* <ALT>-down/<ALT>-up sequence */ > - (wVk != VK_MENU)) /* <ALT>-down...<something else>-up */ > + (VkStripLR(wVk) != VK_MENU)) /* <ALT>-down...<something else>-up */ > message = WM_SYSKEYUP; > TrackSysKey = 0; > } > key_state_table[wVk] &= ~0x80; > + key_state_table[VkStripLR(wVk)] &= ~0x80; > keylp.lp1.previous = 1; > keylp.lp1.transition = 1; > } > @@ -1145,12 +1160,13 @@ > keylp.lp1.transition = 0; > if (!(key_state_table[wVk] & 0x80)) key_state_table[wVk] ^= 0x01; > key_state_table[wVk] |= 0xc0; > + key_state_table[VkStripLR(wVk)] |= 0xc0; > > message = WM_KEYDOWN; > if ((key_state_table[VK_MENU] & 0x80) && !(key_state_table[VK_CONTROL] & 0x80)) > { > message = WM_SYSKEYDOWN; > - TrackSysKey = wVk; > + TrackSysKey = VkStripLR(wVk); It would be much simpler to call VkStripLR() only once and use the result for further processing. -- Dmitry.

1 0